A Retrieval-Augmented Generation skill that enables AI agents to query and retrieve context from SumoNoteBook markdown knowledge bases.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install sumo-notebook-rag
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install sumo-notebook-rag using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Notebook RAG Skill?
The notebook-rag skill is a specialized tool designed for the Openclaw Skills ecosystem to bridge the gap between AI agents and local knowledge repositories. By leveraging vector search technologies, it allows agents to perform semantic queries against a library of markdown files, ensuring that responses are grounded in specific, private technical documentation or personal notes.
Built for developers who require high precision, this Openclaw Skills extension utilizes local embedding models to maintain privacy and speed. It transforms static markdown files into a searchable vector space, making the SumoNoteBook family architecture and learning records instantly accessible to your AI workspace.
Notebook RAG Skill Use Cases
- Retrieving technical architecture details from the SumoNoteBook repository.
- Searching through extensive learning logs and personal engineering notes for specific references.
- Providing grounded context to AI agents during complex coding or research tasks.
- Citing specific markdown files and snippets to validate agent-generated answers.
How Notebook RAG Skill Works
- The user initiates a request using the /notebook-rag command followed by a text query.
- The system generates a 768-dimensional vector embedding of the query using the nomic-embed-text model via Ollama.
- A cosine similarity search is performed against the sumo_notebook table stored in a local LanceDB instance.
- The skill identifies the top 3 most relevant document chunks based on distance scores.
- A formatted response is returned containing file names, relevance scores, and content previews.
Notebook RAG Skill Setup
To integrate this into your Openclaw Skills environment, ensure Ollama is running and the database is initialized:
# Verify Ollama connectivity
curl http://localhost:11434
# Navigate to the script directory
cd C:\butler_sumo\library\SumoNoteBook\scripts
# Perform initial ingestion or a full rebuild of the index
node ingest_notebook.mjs --rebuild
# For future updates, run incremental sync
node ingest_notebook.mjs
Notebook RAG Skill Data Schema & Taxonomy
The skill manages data across a structured vector database and local files:
| Component | Detail |
|---|---|
| Vector DB | LanceDB located at ~/.openclaw/memory/lancedb-pro |
| Table Name | sumo_notebook |
| Index Scope | All .md files in SumoNoteBook (117 files / 260 chunks) |
| Embedding | nomic-embed-text (768 dimensions) |
| Search Metric | Cosine Distance |
Notebook RAG Skill Advanced Features
- Local-first architecture ensuring no data leaves your machine during the RAG process.
- Support for incremental updates to the knowledge base without re-indexing the entire library.
- Integrated similarity scoring to help agents evaluate the reliability of retrieved context.
- Optimized for 768-dimensional embeddings to balance search speed and semantic accuracy.
SKILL.md
Loading
Related Openclaw Skills
Absolute Mode Efficiency Protocol
A protocol designed to maximize token efficiency by stripping all conversational filler and focusing on high-density technical data.
Cati Prova
An automated AI study kit generator that researches educational topics and creates comprehensive learning materials for young students.
Energy Market Intelligence
A specialized AI skill for generating sourced, decision-ready Alberta energy market briefings grounded in AESO data and real-world market structures.
E-Invoice Local Compliance Partner Selection
A sophisticated AI skill designed to systematically identify, filter, and evaluate local e-invoicing partners and compliance channels in international markets.
MySQL Slow Query Analyzer
A professional performance tuning tool that transforms cryptic MySQL EXPLAIN outputs and slow query logs into actionable optimization reports.
Why Skill
A comprehensive knowledge module exploring the etymology, philosophy, and scientific significance of the world's most powerful question.
Featured*
